Are more productive banks always better? · CAFRAL-World Bank 26.02.2020 0/21. Motivation Data &...

Motivation Data & Methodology Empirical Results Interpreting gains Conclusion

Are more productive banks always better?

Rajeswari Sengupta (IGIDR, Mumbai)Harsh Vardhan (SPJIMR Mumbai)

Preliminary. Please do not cite.

5th South Asia Economic Network ConferenceCAFRAL-World Bank

26.02.2020

0 / 21


Contents

1 Motivation

2 Data & Methodology

3 Empirical Results

4 Interpreting gains

5 Conclusion

0 / 21


Productivity in banking

• An important source of economic growth is productivity growth.• Productivity can be studied at the level of the entire economy and also

for specific sectors.• Banking sector in India is the cornerstone of financial intermediation.

I Banks contribute more than 90% of the economy’s commercial credit.I Banking income accounts for roughly 6% of Indian GDP.I With a total asset size of USD 2.2 trillion, it is also among the largest

banking systems in the world.

• We study the productivity growth in Indian banking over the last coupleof decades, across different ownership groups.

• Study of bank efficiency can contribute to policy making and improvingbank management.

1 / 21


Indian Banking: Pre 1991

• Between 1969 and 1995, government owned banks (public sector banksor PSBs) dominated the banking sector, with over 90% share of thebusiness.

• PSBs grew rapidly, not in pursuit of commercial objectives but in line withthe government policy.

• Expanded through large scale recruitment and increased branchnetwork to reach all segments of population.

• Competition was muted and banks had to follow interest rate and pricingregulations.

2 / 21


Indian Banking: 1991-1996

• Deregulation, liberalisation, privatisation reforms ushered in dramaticshift in the banking sector.

• New private banks started operations in 1995-1996.• They started with superior technology platforms that helped them scale

rapidly and efficiently.• PSBs responded by adopting technology (core banking solution etc),

downsizing overstaffed organisations, listing shares publicly etc.

3 / 21


Indian Banking: Mid 1990s-Present

• Competition went up in a more open, and market-oriented environment.• Several structural changes:

I Rapid growth of private and foreign banks,I Expansion of reach of formal banking services,I Reforms to improve income recognition and asset classification,I Use of computers and information technology,I Improved risk management practices,I Emergence of payment banks and small finance banks,I Harmonisation of the regulatory regime with global standards etc.

4 / 21


Indian Banking: Productivity growth

• These changes would affect productivity of the banking sector.• Adoption of technology by PSBs may have resulted in significant

productivity gains.

• One metric for measuring banking productivity is cost of intermediation.• Costs of operations of banking system (costs of manpower,

establishment, IT etc).• Cost of intermediation must come down for the system to enjoy

productivity gains.

5 / 21


Cost of intermediation of the Indian banking system, 1991-2019Ratio of operating costs to assets and income

6 / 21


Cost of intermediation for various groups of banks, 1991-2019Ratio of operating costs to assets and income

1.0

1.5

2.0

2.5

3.0

3.5

4.0

In p

erce

ntag

e te

rms

1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018

Non−SBI PSBsSBI and associates

ForeignPrivate

7 / 21


Research questions

1 What has been the pattern of evolution of productivity growth in theIndian banking sector, during 2002-2018?Productivity grew steadily till 2010 and slowed down thereafter, decliningin the most recent years. Overall, productivity has been stagnant.

2 What has the been the distribution of efficiency gains across variousownership groups of banks?PSBs experienced steep gains during 2002-2010 which got reversedfrom 2015 onwards.Private banks witnessed moderate gains during2002-2010 but productivity stagnation from 2011.

3 What have been the main sources of productivity growth?Frontier shift or technology change has been the main source for PSBswhereas efficiency change or catching up to the frontier has beenimportant for private banks.

8 / 21


Literature

Other countries Berger & Humphrey (1997), Kumbhakar et al. (2001),Canhoto & Dermine (2003), Das & Shanmugam (2004),Sensarma (2006), Casu et al. (2010).

India Bhattacharyya et al. (1997), Sathye (2003), Das et al. (2005),Das & Ghosh (2006), Kumar & Gulati (2009).

Summary: Most studies have explored the effect of deregulation andliberalisation reforms on productivity growth in the banking sector.

9 / 21


Contribution

• Since 2000s many changes have taken place which would impactbanking productivity.

• A study spanning a long time period is able to capture these nuances.• Our methodology allows us to capture efficiency gains over time without

any specific assumption about the production function.

10 / 21


Data envelopment analysis

• Use a linear, input-output programming method called the DataEnvelopment Analysis (DEA).

• Define a non-parametric, efficient production frontier which serves as abenchmark for the measure of efficiency.

• The benchmark is a linear combination of banks included in the sample.• The efficiency scores range from 0 to 1.• Input-oriented approach to efficiency measurement.

11 / 21


Malmquist Index

• DEA cannot be applied to compare banks observed in different periodsof time.

• To analyse productivity growth over time, we use the Malmquist Index(MI)-calculates absolute improvement over time in efficiency.

• Look into sources of productivity gains by separating the efficiencychange into a frontier or technology shift component and an efficiencygain or catching up component.

12 / 21


Intermediation vs. Value added approaches

• DEA and MI use inputs and outputs to calculate efficiency scores.• Intermediation approach

I Inputs: Employee expenses, operating expenses and depositI Output: Loans

• Value added approachI Inputs: Employee expenses, operating expenses and interest expensesI Output: Loans and deposit

• Data: 37 commercial banks; 90% of total banking assets.I 19 PSBs, 11 private banks, 7 foreign banks.

• Sample period: 2002-2018• Source: Prowess database of CMIE.

13 / 21


Summary of DEA results

All banks• 2002: Average efficiency -43.9% and 5.4% banks were on the efficiency

frontier.• 2011: Average efficiency - 87.8% and 21.6% banks were on the

efficiency frontier.• Percentage of efficient banks goes up till 2011 and after that it goes

down.

PSBs vs Private• Intermediation approach: 2008-2015- efficiency scores of PSBs are

higher than private banks; reversed from 2016 onwards.• Value added approach: 2009-2018- efficiency scores of private banks

are higher

14 / 21


Summary of MI resultsAverage efficiency gains for all banks: Intermediation approach

Periods MI TC EC2003-2006 1.014 0.892 1.1772007-2010 1.054 1.001 1.0562011-2014 1.024 1.038 0.9872015-2018 0.946 0.969 0.9872003-2010 1.034 0.946 1.1172011-2018 0.985 1.003 0.9872002-2018 1.010 0.975 1.052

1 2003-2010: An increase in average efficiency of 3.4%.

2 2011-2018: A decrease in average efficiency of 2.2%.

3 2002-2018: Average efficiency improvement only by 1%.

4 Efficiency change (‘catching-up’) more important than technologychange.

15 / 21


Efficiency gains for bank groups: Intermediation approach

• 2003-2010: Efficiency of PSBs improved by 5.9%, and for private banks,by 1.4%.

I For PSBs, gains from frontier shift (technological change) were greater thangains catching up effect.

• 2011-2018: Efficiency of PSBs declined by 2.3%; for private banks, itwas stagnant at 0.2%.

• 2002-2018: Average efficiency score for PSBs larger than private banksindicating superior improvement in productivity.

16 / 21


Top 5 banks by efficiency gains, 2002-2018

PSBs PrivateAllahabad Bank H D F C Bank Ltd.

Canara Bank Karnataka Bank Ltd.Oriental Bank Of Commerce Federal Bank Ltd.

State Bank Of India Axis Bank Ltd.Indian Bank South Indian Bank Ltd.

17 / 21


How to interpret efficiency gains in banking?

• Conventional measurement: Indian banking sector witnessed significantproductivity gains till about 2010.

• An important output is credit extended by banks.• If a bank gives out a higher volume of loans, this shows up as a rise in

productivity.• Irrespective of the quality of credit extended by the bank.• Consequences would be seen a few years later when much of the loans

go bad.• Periods of rapid credit growth will show strong productivity gains but

could be periods where credit quality goes down.

18 / 21


Productivity growth followed by worsening bank balance sheet

• 2003-2010: Bank credit grew by 25%, especially for PSBs.• 2002-2010: Steady productivity growth.• Post 2010: NPAs in the banking sector began rising.• NPAs peaked in 2018 at 11.1% for the banking system and 14.6% for

PSBs.

• Raises questions about how we measure productivity and how weinterpret the results.

19 / 21


Conclusion

• Indian banking sector experienced steady productivity growth, PSBsmore than the private banks, during 2002- 2010.

• Technology change (frontier shift) was more important for PSBs whereasprivate banks benefitted from faster efficiency change (catching up to thefrontier).

• 2011 onwards productivity growth slowed down dramatically.• Decline in productivity from 2015 onwards, led by PSBs.• 2002-2018: no significant growth in productivity.

I A new set of metrics that factor in the unique nature of banking businessmay be needed.

20 / 21


Thank [email protected]

[email protected]

21 / 21

Are more productive banks always better? · CAFRAL-World Bank 26.02.2020 0/21. Motivation Data &...

Documents

Transcript of Are more productive banks always better? · CAFRAL-World Bank 26.02.2020 0/21. Motivation Data &...